| Summary: | 碩士 === 輔仁大學 === 資訊管理學系碩士在職專班 === 105 === This study explores the applicability of money management and risk control of trading strategy of the Taiwan Stock Exchange Capitalization Weighted Stock Index futures (TAIEX Futures) in the Swing Trading. The goal is to develop a trading strategy model with dynamic decision and money management. This study uses the Gene Expression Programming (GEP) to optimize the search and planning capabilities of the application of program trading. Through the multi-gene chromosome coding and random numerical constants (RNCs), this study innovatively adopting GEP to get the volatility degree of dynamic days of technical indicators in different periods by which we measure the relative strength of indicators and avoid the target passivation to provide more market-sensitive of the transaction signals. GEP decision trees (GEP-DT) are planned for the genes management based on market and transaction performance data. With the dynamic evaluation function (Dynamic evaluation function) concept, various money management methods are applied to improve the applicability problem generated by the over-reliance on a single money management method in Swing Trading. Finally, supplemented by the addition and subtraction strategy and the stop-loss mechanism of the capital allocation, we expect to develop futures trading investment strategies which bring better risk premium and adaptability.
Experimental results show that the Fixed Fraction method has the best overall profit and risk-control capability in Swing Trading. The Fixed Fraction method has the best overall performance and it is suggested for all investors. However, in other markets, it is suitable for the risk-neutral investors. Kelly formula, with the best profit and adaptability in market correction period, is suitable for risk takers due to high MDD. Optimal F is suitable for risk avoiders due to its medium profit performance, adaptability to all kinds of swing trading markets and capability of risk control; Martingale and Anti-Martingale strategy with the highest MDD and the easy-to-oscillate profitability, is suitable for risk takers. Comparing with no overweight-underweight strategy, considering the chip indicator and overweight or underweight timely helps to increase the profit performance and avoid losses. With stop-loss mechanism, the risk premium is maintained steadily; without it, it is difficult to be kept.
|
|---|
